• Alexander, M. A., , U. S. Bhatt, , J. E. Walsh, , M. S. Timlin, , J. S. Miller, , and J. D. Scott, 2004: The atmospheric response to realistic Arctic sea ice anomalies in an AGCM during winter. J. Climate, 17, 890905, doi:10.1175/1520-0442(2004)017<0890:TARTRA>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Brodzik, M., , and R. Armstrong, 2013: Northern Hemisphere EASE-Grid 2.0 weekly snow cover and sea ice extent version 4. National Snow and Ice Data Center, Boulder, CO, digital media. [Available online at https://nsidc.org/data/docs/daac/nsidc0046_nh_ease_snow_seaice.gd.html.]

  • Chen, Z., , R. Wu, , and W. Chen, 2014: Distinguishing interannual variations of the northern and southern modes of the East Asian winter monsoon. J. Climate, 27, 835851, doi:10.1175/JCLI-D-13-00314.1.

    • Search Google Scholar
    • Export Citation
  • Comiso, J. C., , C. L. Parkinson, , R. Gersten, , and L. Stock, 2008: Accelerated decline in the Arctic sea ice cover. Geophys. Res. Lett., 35, L01703, doi:10.1029/2007GL031972.

    • Search Google Scholar
    • Export Citation
  • Deser, C., , G. Magnusdottir, , R. Saravanan, , and A. Philips, 2004: The effects of North Atlantic SST and sea ice anomalies on the winter circulation in CCM3. Part II: Direct and indirect components of the response. J. Climate, 17, 877889, doi:10.1175/1520-0442(2004)017<0877:TEONAS>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Francis, J. A., , W. Chan, , D. J. Leathers, , J. R. Miller, , and D. E. Veron, 2009: Winter Northern Hemisphere weather patterns remember summer Arctic sea-ice extent. Geophys. Res. Lett., 36, L07503, doi:10.1029/2009GL037274.

    • Search Google Scholar
    • Export Citation
  • Gong, G., , D. Entekhabi, , and J. Cohen, 2003: Modeled Northern Hemisphere winter climate response to realistic snow anomalies. J. Climate, 16, 39173931, doi:10.1175/1520-0442(2003)016<3917:MNHWCR>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • He, S.-P., , and H.-J. Wang, 2013: An integrated East Asian winter monsoon index and its interannual variability (in Chinese). Chin. J. Atmos. Sci., 36 (3), 523538.

    • Search Google Scholar
    • Export Citation
  • Honda, M., , J. Inoue, , and S. Yamane, 2009: Influence of low Arctic sea-ice minima on anomalously cold Eurasian winters. Geophys. Res. Lett., 36, L08707, doi:10.1029/2008GL037079.

    • Search Google Scholar
    • Export Citation
  • Hopsch, S., , J. Cohen, , and K. Dethloff, 2012: Analysis of a link between fall Arctic sea ice concentration and atmospheric patterns in the following winter. Tellus, 64A, 18624, doi:10.3402/tellusa.v64i0.18624.

    • Search Google Scholar
    • Export Citation
  • Inoue, J., , M. E. Hori, , and K. Takaya, 2012: The role of Barents Sea ice in the wintertime cyclone track and emergence of a warm-Arctic cold-Siberian anomaly. J. Climate, 25, 25612568, doi:10.1175/JCLI-D-11-00449.1.

    • Search Google Scholar
    • Export Citation
  • Jaiser, R., , K. Dethloff, , D. Handorf, , A. Rinke, , and J. Cohen, 2012: Impact of sea ice cover changes on the Northern Hemisphere atmospheric winter circulation. Tellus, 64A, 11595, doi:10.3402/tellusa.v64i0.11595.

    • Search Google Scholar
    • Export Citation
  • Jhun, J.-G., , and E.-J. Lee, 2004: A new East Asian winter monsoon index and associated characteristics of the winter monsoon. J. Climate, 17, 711726, doi:10.1175/1520-0442(2004)017<0711:ANEAWM>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Kalnay, E., and Coauthors, 1996: The NCEP/NCAR 40-Year Reanalysis Project. Bull. Amer. Meteor. Soc., 77, 437471, doi:10.1175/1520-0477(1996)077<0437:TNYRP>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Kang, L., , W. Chen, , L. Wang, , and L. Chen, 2009: Interannual variations of winter temperature in China and their relationship with the atmospheric circulation and sea surface temperature (in Chinese). Climate Environ. Res., 14 (1), 4553.

    • Search Google Scholar
    • Export Citation
  • Li, F., , and H.-J. Wang, 2013: Relationship between Bering Sea ice cover and East Asian winter monsoon year-to-year variations. Adv. Atmos. Sci., 30, 4856, doi:10.1007/s00376-012-2071-2.

    • Search Google Scholar
    • Export Citation
  • Liu, G., , L.-R. Ji, , S.-Q. Sun, , and Y.-F. Xin, 2012a: Low- and mid-high latitude components of the East Asian winter monsoon and their reflecting variations in winter climate over eastern China. Atmos. Oceanic Sci. Lett., 5, 195200.

    • Search Google Scholar
    • Export Citation
  • Liu, J., , J. A. Curry, , H.-J. Wang, , M. R. Song, , and R. M. Horton, 2012b: Impact of declining Arctic sea ice on winter snowfall. Proc. Natl. Acad. Sci. USA, 109, 40744079, doi:10.1073/pnas.1114910109.

    • Search Google Scholar
    • Export Citation
  • Orsolini, Y. J., , R. Senan, , R. E. Benestad, , and A. Melsom, 2012: Autumn atmospheric response to the 2007 low Arctic sea ice extent in coupled ocean-atmosphere hindcasts. Climate Dyn., 38, 24372448, doi:10.1007/s00382-011-1169-z.

    • Search Google Scholar
    • Export Citation
  • Rayner, N. A., , D. E. Parker, , E. B. Horton, , C. K. Folland, , L. V. Alexander, , D. P. Rowell, , E. C. Kent, , and A. Kaplan, 2003: Global analyses of sea surface temperature, sea ice, and night marine air temperature since the late nineteenth century. J. Geophys. Res., 108, 4407, doi:10.1029/2002JD002670.

    • Search Google Scholar
    • Export Citation
  • Smith, T. M., , R. W. Reynolds, , T. C. Peterson, , and J. Lawrimore, 2008: Improvements to NOAA’s historical merged land–ocean surface temperature analysis (1880–2006). J. Climate, 21, 22832296, doi:10.1175/2007JCLI2100.1.

    • Search Google Scholar
    • Export Citation
  • Wang, B., , Z. Wu, , C. P. Chang, , J. Liu, , J. Li, , and T. Zhou, 2010: Another look at interannual-to-interdecadal variations of the East Asian winter monsoon: The northern and southern temperature modes. J. Climate, 23, 14951512, doi:10.1175/2009JCLI3243.1.

    • Search Google Scholar
    • Export Citation
  • Wang, L., , and W. Chen, 2010: How well do existing indices measure the strength of the East Asian winter monsoon? Adv. Atmos. Sci., 27, 855870, doi:10.1007/s00376-009-9094-3.

    • Search Google Scholar
    • Export Citation
  • Watanabe, M., , and T. Nitta, 1999: Decadal change in the atmospheric circulation and associated surface climate variations in the Northern Hemispheric winter. J. Climate, 12, 494510, doi:10.1175/1520-0442(1999)012<0494:DCITAC>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Wu, B.-Y., , and J. Wang, 2002: Winter Arctic Oscillation, Siberian High and East Asian winter monsoon. Geophys. Res. Lett., 29, 1897, doi:10.1029/2002GL015373.

    • Search Google Scholar
    • Export Citation
  • Wu, B.-Y., , R.-H. Huang, , and D.-Y. Gao, 1999: Impact of variations of winter sea-ice extents in the Kara/Barents Seas on winter monsoon over East Asia (in Chinese). Acta Meteor. Sin., 13, 141153.

    • Search Google Scholar
    • Export Citation
  • Wu, B.-Y., , R.-H. Zhang, , and R. Arrigo, 2006: Distinct modes of the East Asian winter monsoon. Mon. Wea. Rev., 134, 21652179, doi:10.1175/MWR3150.1.

    • Search Google Scholar
    • Export Citation
  • Wu, B.-Y., , J.-Z. Su, , and R.-H. Zhang, 2011: Effects of autumn-winter Arctic sea ice on winter Siberian High. Chin. Sci. Bull., 56, 32203228, doi:10.1007/s11434-011-4696-4.

    • Search Google Scholar
    • Export Citation
  • View in gallery

    (a) Climatological mean and (b) standard deviation of autumn mean sea ice concentration based on the period 1979–2011. The outer latitude circle is 60°N.

  • View in gallery

    (a) Standard deviations of the interannual component of autumn mean sea ice concentration based on the period 1979–2011 and (b) the percentage of variance of the autumn mean sea ice concentration explained by the interannual components. The outer latitude circle is 60°N.

  • View in gallery

    Correlation coefficients between autumn mean sea ice concentration and the (a) N index and (b) S index of EAWM based on the period 1979–2011. The two domains enclosed by thick lines in (a) denote the regions that are used to define the WSIC and ESIC. The outer latitude circle is 60°N.

  • View in gallery

    The normalized (a) original and (b) interannual (b) time series of the WSIC (solid line) and ESIC (dashed line) for the period 1979–2011.

  • View in gallery

    Correlation coefficient of (a),(b) autumn and (c),(d) winter sea ice concentration with autumn (left) WSIC and (right) ESIC for the period 1979–2011. The two sectors (thick solid lines) denote the regions used to define the WSIC and ESIC.

  • View in gallery

    Autumn air temperature anomalies at sigma level 0.995 obtained by regression on autumn (a) WSIC and (b) ESIC for the period 1979–2011. The contour interval is 0.2°C. The shaded areas denote anomalies that are significant at the 95% confidence level according to the Student’s t test. The outer latitude circle is 20°N.

  • View in gallery

    (a),(b) Autumn SLP and (c),(d) 500-hPa geopotential height and corresponding wave activity flux anomalies obtained by regression on autumn (a),(c) WSIC and (b),(d) ESIC for the period 1979–2011. The contour interval is 0.3 hPa in (a),(b) and 4 m in (c),(d). The shaded areas denote anomalies significant at the 95% confidence level according to a Student’s t test. The scale for the wave activity flux is shown at the top right of the panels. The outer latitude circle is 20°N.

  • View in gallery

    Profiles of autumn temperature anomalies (°C) averaged over the WSIC domain (solid) and ESIC domain (dashed) obtained by regression on the WSIC and ESIC index for the period 1979–2011.

  • View in gallery

    As in Fig. 6, but for winter air temperature anomalies.

  • View in gallery

    As in Fig. 7, but for (a),(b) winter SLP and (c),(d) 500-hPa geopotential height and correspondent wave activity flux anomalies.

  • View in gallery

    Height–latitude cross sections of winter temperature anomalies (°C) averaged over the longitudinal band of 60°–120°E obtained by regression on the (a) WSIC and (b) ESIC for the period 1979–2011. The contour interval is 0.1°C and the zero line is omitted.

  • View in gallery

    As in Fig. 11, but for winter geopotential height anomalies with the contour interval of 3 m.

  • View in gallery

    As in Fig. 11, but for winter zonal wind anomalies with the contour interval of 0.3 m s−1.

  • View in gallery

    As in Fig. 6, but for winter SST anomalies. The contour interval is 0.1°C. The outer latitude circle is 0°.

  • View in gallery

    As in Fig. 6, but for original winter SST anomalies. The contour interval is 0.1°C. The outer latitude circle is 0°.

  • View in gallery

    Winter snow cover frequency anomalies obtained by regression on autumn (a) WSIC and (b) ESIC for the period 1979–2010.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 48 48 10
PDF Downloads 44 44 6

Impacts of Autumn Arctic Sea Ice Concentration Changes on the East Asian Winter Monsoon Variability

View More View Less
  • 1 Center for Monsoon System Research, Institute of Atmospheric Physics, Chinese Academy of Sciences, and University of Chinese Academy of Sciences, Beijing, China
  • 2 Institute of Space and Earth Information Science and Shenzhen Research Institute, Chinese University of Hong Kong, Hong Kong, China
  • 3 Center for Monsoon System Research, Institute of Atmospheric Physics, Chinese Academy of Sciences, Beijing, China
© Get Permissions
Full access

Abstract

The present study investigated the impacts of autumn Arctic sea ice concentration (SIC) changes on the East Asian winter monsoon (EAWM) and associated climate and circulation on the interannual time scale. It is found that the Arctic SIC anomalies have little impact on the southern mode of EAWM, but the northern mode is significantly associated with both western and eastern Arctic SIC anomalies. When there is less (more) SIC in eastern (western) Arctic, the EAWM tends to be stronger. The concurrent surface air temperature anomalies are induced both locally due to the direct effect of ice cover and in remote regions through anomalous wind advection. Analysis showed that eastern Arctic SIC anomalies have a larger effect on local atmospheric stability of the lower troposphere than western Arctic SIC anomalies. Winter temperature over the midlatitudes of East Asia is lower when there is more (less) SIC in the western (eastern) Arctic. The atmospheric response to the Arctic SIC anomalies is dominantly barotropic in autumn, and changes to baroclinic over the midlatitudes of Asia, but remains barotropic in other regions in winter. The mid- to high-latitude circulation systems, including the Siberian high, the East Asian trough, and the East Asian westerly jet stream, play important roles in connecting autumn Arctic SIC anomalies and the northern mode of the EAWM variability. No obvious concurrent sea surface temperature anomalies accompany Arctic SIC variations on the interannual time scale, indicating that the Arctic SIC anomalies have independent impacts on the East Asian winter climate.

Corresponding author address: Renguang Wu, Fok Ying Tung Remote Sensing Science Building, Chinese University of Hong Kong, Shatin, NT, Hong Kong, China. E-mail: renguang@cuhk.edu.hk

Abstract

The present study investigated the impacts of autumn Arctic sea ice concentration (SIC) changes on the East Asian winter monsoon (EAWM) and associated climate and circulation on the interannual time scale. It is found that the Arctic SIC anomalies have little impact on the southern mode of EAWM, but the northern mode is significantly associated with both western and eastern Arctic SIC anomalies. When there is less (more) SIC in eastern (western) Arctic, the EAWM tends to be stronger. The concurrent surface air temperature anomalies are induced both locally due to the direct effect of ice cover and in remote regions through anomalous wind advection. Analysis showed that eastern Arctic SIC anomalies have a larger effect on local atmospheric stability of the lower troposphere than western Arctic SIC anomalies. Winter temperature over the midlatitudes of East Asia is lower when there is more (less) SIC in the western (eastern) Arctic. The atmospheric response to the Arctic SIC anomalies is dominantly barotropic in autumn, and changes to baroclinic over the midlatitudes of Asia, but remains barotropic in other regions in winter. The mid- to high-latitude circulation systems, including the Siberian high, the East Asian trough, and the East Asian westerly jet stream, play important roles in connecting autumn Arctic SIC anomalies and the northern mode of the EAWM variability. No obvious concurrent sea surface temperature anomalies accompany Arctic SIC variations on the interannual time scale, indicating that the Arctic SIC anomalies have independent impacts on the East Asian winter climate.

Corresponding author address: Renguang Wu, Fok Ying Tung Remote Sensing Science Building, Chinese University of Hong Kong, Shatin, NT, Hong Kong, China. E-mail: renguang@cuhk.edu.hk

1. Introduction

As one of the most active systems in the Northern Hemisphere during boreal winter, the East Asian winter monsoon (EAWM) exerts large influences on the Asian–Pacific winter climate. The climate anomalies associated with the EAWM variability show obvious year-to-year variations in their meridional coverage and intensity as well as in the causes (e.g., Wang and Chen 2010). Coherent and distinct winter temperature anomalies have been observed between the mid- to high latitudes and low latitudes of East Asia in different years. This leads to the difficulty of using a single index to fully describe the EAWM variability and associated climate anomalies, and to unravel the factors. Thus, it is necessary to distinguish the northern and southern components of the EAWM variability (e.g., Wu et al. 2006; Wang et al. 2010; Liu et al. 2012a; Chen et al. 2014).

Studies have revealed the presence of two distinct modes in climate and circulation associated with the EAWM variability. Kang et al. (2009) identified two modes of winter surface air temperature variations in China. Wang et al. (2010) extracted a northern mode and a southern mode of EAWM variability based on wintertime surface air temperature anomalies over Asia and the western Pacific. Liu et al. (2012a) defined two wind indices for the EAWM variability: the low-latitude EAWM index and the mid- to high-latitude EAWM index. Using composite analysis, Chen et al. (2014) identified individual and coherent features related to the two modes of EAWM variability.

Previous studies have identified different factors for the two modes of the EAWM variability. The southern EAWM mode is closely related to El Niño–Southern Oscillation (ENSO) (e.g., Wang et al. 2010; Chen et al. 2014). The northern EAWM mode does not show an obvious relationship to tropical sea surface temperature (SST) as well as the Arctic Oscillation (AO) on the interannual time scale (Chen et al. 2014), but has a close relationship with the AO on the interdecadal time scale (e.g., He and Wang 2013). As the factors for the EAWM variability depend upon the time scale (Wang et al. 2010; Wang and Chen 2010), Chen et al. (2014) pointed out that it is necessary to separate the interannual and interdecadal components of the EAWM variability for a better understanding of the causes of the EAWM-related climate anomalies.

The EAWM variability has been linked to variations in snow cover over the Eurasian continent (e.g., Watanabe and Nitta 1999; Gong et al. 2003; Jhun and Lee 2004; Wang et al. 2010; Chen et al. 2014) and Arctic sea ice (e.g., Wu et al. 1999; Honda et al. 2009; Wu et al. 2011; Li and Wang 2013; Chen et al. 2014). Wang et al. (2010) indicated that the cold state of the northern temperature mode of EAWM is preceded by excessive autumn snow cover over southern Siberia–Mongolia, whereas the cold state of the southern temperature mode is preceded by reduced snow cover over northeast Siberia. Chen et al. (2014) showed that, on the interannual time scale, distinct winter snow and sea ice anomalies appear as responses to wind and surface temperature changes associated with the two modes of the EAWM variability, but it is unclear how the snow and sea ice anomalies affect variability of the two modes. The present study aims to investigate the impacts of Arctic sea ice anomalies on the two modes of the EAWM variability and to identify predictors of the two modes in the sea ice field.

Sea ice is an important component in the global climate system since it regulates the heat exchange between the ocean and atmosphere. Changes in the Arctic sea ice influence local as well as the global climate through modulating surface heat exchange and atmospheric circulation (e.g., Alexander et al. 2004; Deser et al. 2004; Honda et al. 2009). Several studies have investigated the responses of the atmosphere to late summer–fall or winter Arctic sea ice anomalies (Honda et al. 2009; Francis et al. 2009; Wu et al. 1999, 2011; Hopsch et al. 2012; Inoue et al. 2012; Jaiser et al. 2012; Liu et al. 2012b; Li and Wang 2013). Wu et al. (1999) showed that a winter heavy sea ice in the Barents–Kara Seas is accompanied by a weakened EAWM. Honda et al. (2009) found that a decrease in the Arctic sea ice during fall is followed by significant cold anomalies over the Far East in early winter and cold anomalies from Europe to the Far East in late winter. Francis et al. (2009) indicated that the Aleutian and Icelandic lows and the polar jet stream are weaker during autumn and winter following less Arctic sea ice in September. Wu et al. (2011) indicated that the intensity of the winter Siberian high is negatively correlated with the autumn–winter Arctic sea ice concentration in the eastern Arctic and Eurasian marginal seas and September sea ice concentration provides a potential precursor for the winter Siberian high. Li and Wang (2013) pointed out that the EAWM variability is tightly related to the Bering Sea ice cover variation with light sea ice cover accompanied by a stronger EAWM.

The above studies, however, did not investigate the association between the Arctic sea ice and the two modes of the EAWM variability separately. In addition, they mostly focused on the impacts of sea ice anomalies in the eastern Arctic Ocean (e.g., the Barents–Kara Seas and Siberian coast seas) and the roles of sea ice changes in the western Arctic Ocean have received relatively little attention. The present study investigates the signals of sea ice in both eastern and western Arctic regions in the two modes of the EAWM variability and their plausible connections. Several studies considered both actual and detrended sea ice anomalies (e.g., Francis et al. 2009; Honda et al. 2009; Hopsch et al. 2012). Some studies were based on detrended sea ice variations (e.g., Wu et al. 2011; Liu et al. 2012b). Arctic sea ice has experienced an accelerated decline starting in the late 1990s (Comiso et al. 2008), and there is no or only a weak trend before the 1990s. Thus, it is inappropriate to compute a linear trend for the entire time series. The time series with the linear trend removed may include variations on both interannual and decadal time scales. Distinguishing from previous studies, the present study focuses on the effects of the interannual variations in the Arctic sea ice concentration.

The rest of the text is organized as follows. The datasets and methods used in the present study are described in section 2. Section 3 introduces the variability of autumn Arctic sea ice concentration. The definition of the Arctic sea ice concentration indices is presented in section 4. The impacts of the Arctic sea ice concentration anomalies on autumn and winter climate and circulation are addressed in section 5. Section 6 includes a summary and discussion.

2. Data and methods

The sea ice concentration (SIC) data used in this study were obtained from the Met Office Hadley Centre with a resolution of 1° × 1° available since 1870 (Rayner et al. 2003). The SST data gridded at 2° × 2° resolution used in this study were taken from the National Oceanic and Atmospheric Administration extended reconstructed SST version 3b (ERSST V3b) (Smith et al. 2008). The SST dataset is available after 1854. We used snow cover data (version 4) provided by the National Snow and Ice Data Center (NSIDC) (Brodzik and Armstrong 2013). The original snow cover is at weekly intervals for the period 3 October 1966–31 December 2010. We have converted the raw snow cover data to regular 1° × 1° grid for our analysis.

Monthly mean sea level pressure (SLP), wind, temperature, and geopotential height fields from the National Centers for Environmental Prediction–National Center for Atmospheric Research (NCEP–NCAR) reanalysis (Kalnay et al. 1996) data are available after 1948. The NCEP–NCAR reanalysis data are gridded at 2.5° × 2.5° resolution except for surface variables that have a T62 spectral resolution of approximately 1.9° × 1.9°.

The time period analyzed in this study is from 1979 to 2012. We focus on the autumn [September–November (SON)] and winter [December–February (DJF)] atmospheric response to the autumn Arctic sea ice anomalies and the 1979 winter refers to the three months from December 1979 to February 1980. In the present study, we focus on the interannual time scale and a harmonic analysis has been applied to all of the variables to exclude the variations with periods longer than 9 yr. Following Chen et al. (2014), we used the area-mean 1000-hPa meridional wind anomalies in DJF over the regions 10°–25°N, 105°–135°E and 35°–55°N, 110°–125°E to represent the southern and northern modes of the EAWM variability, respectively, which are denoted as the S index and N index. The above regions are selected based on the distribution of loading of the leading modes and standard deviation of 1000-hPa meridional wind anomalies in DJF. Large loadings and standard deviations are identified in the above two regions. The two modes are used to distinguish the difference in the EAWM variability between low latitudes and mid- to high latitudes. Analysis shows that the northern mode is related to circulation systems over the mid- to high latitudes, such as the Siberian high, the Aleutian low, the East Asian trough, and the East Asian westerly jet stream, but not to tropical circulation systems, whereas the southern mode is closely associated with global tropical circulation, but not with Eurasian circulation (Chen et al. 2014).

3. Variability of autumn Arctic sea ice concentration

The climatologically mean autumn SIC shows the largest values in the polar region, and the value decreases southward with relatively small SIC in the marginal seas where annual sea ice retreat occurs (Fig. 1a). The corresponding standard deviation of autumn SIC is small around the North Pole where there is perennial sea ice cover and it increases southward to regions where a large gradient in mean concentration is observed (Fig. 1b). The largest standard deviations appear in the eastern Arctic, from the northern Barents Sea to the eastern Siberian Sea and the Beaufort Sea. These are the regions where a declining trend in autumn SIC is significant after 1990 (Wu et al. 2011). Thus, it is expected that the trend has a large contribution to the standard deviation in the above regions.

Fig. 1.
Fig. 1.

(a) Climatological mean and (b) standard deviation of autumn mean sea ice concentration based on the period 1979–2011. The outer latitude circle is 60°N.

Citation: Journal of Climate 27, 14; 10.1175/JCLI-D-13-00731.1

Since the present study focuses on the interannual time scale, we calculate the standard deviation of the interannual component of autumn SIC variations (Fig. 2a). The spatial pattern of interannual standard deviation is similar to that in Fig. 1b except that the values are smaller. Relatively large values are located in the Arctic marginal seas. The percentage of the autumn mean SIC variance explained by the interannual component, which is shown in Fig. 2b, is between 0.4 and 0.8 in most regions except for the eastern Siberian Sea and the Beaufort Sea. This indicates that the interannual variability has nearly equal or more importance compared to the slow variability (including both trend and interdecadal variability) in these regions. In the eastern Siberian Sea and the Beaufort Sea where the largest linear trend is observed (Wu et al. 2011), however, the trend and the interdecadal variability is dominant.

Fig. 2.
Fig. 2.

(a) Standard deviations of the interannual component of autumn mean sea ice concentration based on the period 1979–2011 and (b) the percentage of variance of the autumn mean sea ice concentration explained by the interannual components. The outer latitude circle is 60°N.

Citation: Journal of Climate 27, 14; 10.1175/JCLI-D-13-00731.1

In the following analysis, we focus on interannual variations of the autumn Arctic SIC and their impacts. The slow variations with periods longer than 9 yr have been excluded except when it is noted. This is due to two reasons. One is that the contribution of the interannual and slow variations to the total sea ice variance displays a different spatial distribution as shown in Fig. 2b. The other, which is more important, is that the relationship between sea ice and EAWM-related climate and circulation may differ between the interannual and interdecadal time scales.

4. Definition of the Arctic sea ice concentration indices

To explore the relationship between autumn Arctic SIC anomalies and the two modes of the EAWM variability, a correlation analysis was performed for the N index and S index separately. The N index displays a negative correlation with sea ice from the Kara Sea to the eastern Siberian Sea and a positive correlation with sea ice in the western Arctic (Fig. 3a). Note that the standard deviation is larger in the former region than in the latter region whether for total or interannual anomalies (Figs. 1b and 2a). The S index shows a weak correlation in most regions except in the Kara Sea, the Beaufort Sea, around the North Pole, and in coastal regions of Canada (Fig. 3b). Note that the correlation in the northern Barents Sea is weak for both N and S indices. In comparison, the correlation distribution is more organized and significant for the N index than for the S index. Since the southern mode of EAWM does not show an obvious organized relationship to the Arctic sea ice on the interannual time scale, we will only investigate the effects of the Arctic sea ice on the northern mode of EAWM in the present study. According to Fig. 3a, we define area-mean SON SIC anomalies over the regions 75.5°–85.5°N, 160°W–75°E and 78.5°–87.5°N, 30°–150°W, denoted as the ESIC and WSIC, to represent the eastern and western Arctic SIC variations, respectively. The correlation coefficient between the N index and the WSIC is 0.48, and that between the N index and the ESIC is −0.49, both exceeding the 95% confidence level according to the Student’s t test. This indicates that our SIC indices are appropriate to study the effects of autumn Arctic SIC on the northern mode of the EAWM variability. Note that the standard deviation of SIC anomalies are much larger in the eastern Arctic than in the western Arctic (Fig. 2a).

Fig. 3.
Fig. 3.

Correlation coefficients between autumn mean sea ice concentration and the (a) N index and (b) S index of EAWM based on the period 1979–2011. The two domains enclosed by thick lines in (a) denote the regions that are used to define the WSIC and ESIC. The outer latitude circle is 60°N.

Citation: Journal of Climate 27, 14; 10.1175/JCLI-D-13-00731.1

For comparison, we have examined the correlation of winter SIC anomalies with the N index and the S index. The correlation with the N index appears weaker and less organized compared to autumn. The correlation with the S index differs from that in autumn, indicative of weak persistence of Arctic SIC anomalies from autumn to winter. The correlation coefficient of winter WSIC and ESIC with the N index is 0.27 and −0.05, respectively, and that with the S index is 0.05 and −0.03, respectively, all of which are not significant.

Somewhat different domains have been used in previous studies in defining sea ice indices to study the relationship between the Arctic sea ice and climate. Honda et al. (2009) defined the September sea ice index using area mean SIC averaged in the region of 72°–82°N, 30°E–180° from the Barents Sea to the East Siberian Sea. Wu et al. (2011) used regional-averaged September SIC in the region of 76.5°–83.5°N, 60.5°–149.5°E as an index. Jaiser et al. (2012) calculated an area mean SIC in the domain of 75°–83°N, 60°E–180° during August and September. Hopsch et al. (2012) used the SIC averaged in the region of 70°–85°N, 90°E–150°W in September. The domains used in defining the sea ice indices are chosen either based on the spatial distribution of large sea ice variability (Honda et al. 2009; Jaiser et al. 2012; Hopsch et al. 2012) or based on the region of significant correlation with the winter Siberian high intensity (Wu et al. 2011). The domains used in these previous studies are similar and are mainly located in the eastern Arctic, from the Kara Sea to eastern Siberian Sea. In the present study, we choose the regions based on the distribution of significant correlation between the northern mode of EAWM and the autumn Arctic SIC. The ESIC is similar to the aforementioned indices, while the WSIC is proposed for the first time.

The interannual standard deviation of WSIC and ESIC is about 0.01 and 0.05, respectively. Figure 4 shows the autumn WSIC and ESIC indices (Fig. 4a) and their interannual components (Fig. 4b) normalized using respective standard deviations. As shown in Fig. 4a, the ESIC time series displays an overall decrease during the analysis period, which agrees with previous studies (e.g., Francis et al. 2009; Hopsch et al. 2012). However, the decline is not linear. The ESIC remained above zero with some fluctuations in the 1980s and early 1990s and started to decrease after the mid-1990s, first slower and then faster. The WSIC time series shows no significant trend but some decadal variability. The WSIC is below zero before 1987, above zero in most years during 1987–97, and below zero again in most years during 1998–2008. Since there is no obvious or coherent trend during 1979–2011 for both ESIC and WSIC, it is improper to apply a linear trend removal to the entire time period as done by previous studies. In addition, the linear-trend-removed time series may include both interannual and interdecadal variations. The two indices exhibit pronounced interannual variations, especially in the recent decades (Fig. 4b). The correlation coefficient between the two indices is −0.23, which is not significant at the 90% confidence level. This indicates that the interannual variations of the ESIC and WSIC indices tend to be independent. Nevertheless, the two indices display clearly opposite and large anomalies during 2005–08 (Fig. 4b).

Fig. 4.
Fig. 4.

The normalized (a) original and (b) interannual (b) time series of the WSIC (solid line) and ESIC (dashed line) for the period 1979–2011.

Citation: Journal of Climate 27, 14; 10.1175/JCLI-D-13-00731.1

In Fig. 5, we present the correlation of the autumn and winter Arctic SIC with the autumn WSIC and ESIC on the interannual time scale. In autumn, there are high positive correlations in the regions used in defining the SIC indices, and weak correlations are seen in the eastern Arctic for WSIC and in the western Arctic for ESIC, respectively (Figs. 5a,b). These results confirm that the two SIC indices represent concurrent SIC variations in the corresponding domains well and the SIC variations in the eastern Arctic and western Arctic tend to be independent of each other. The autumn WSIC is still significantly correlated with the winter SIC in the western Arctic (Fig. 5c). The correlation coefficient between autumn and winter WSIC is 0.52, exceeding the 95% confidence level. This indicates that the western Arctic SIC anomalies tend to persist from autumn to winter. On the contrary, there is weak correlation between the autumn ESIC and winter SIC in the eastern Arctic (Fig. 5d). The correlation coefficient between autumn and winter ESIC is only 0.18. This signifies that eastern Arctic SIC anomalies have a weak persistence from autumn to winter.

Fig. 5.
Fig. 5.

Correlation coefficient of (a),(b) autumn and (c),(d) winter sea ice concentration with autumn (left) WSIC and (right) ESIC for the period 1979–2011. The two sectors (thick solid lines) denote the regions used to define the WSIC and ESIC.

Citation: Journal of Climate 27, 14; 10.1175/JCLI-D-13-00731.1

Based on an analysis of linear-trend-removed data, Wu et al. (2011) obtained that the regional-averaged (76.5°–83.5°N, 60.5°–149.5°E) September SIC was significantly correlated with the winter SIC in the Barents–Kara Seas, implying a persistence of the SIC anomalies from autumn to winter. However, there is no obvious correlation between ESIC and the SIC in the Barents–Kara Seas on the interannual time scale (Figs. 5b,d). This discrepancy suggests that the persistence of sea ice anomalies in Wu et al. (2011) is contributed from the interdecadal SIC anomalies. To confirm this, we have calculated the correlations between the autumn/winter Arctic SIC anomalies and the autumn WSIC and ESIC indices without time filtering. The results show that the autumn ESIC has a significant correlation with the SIC in the Barents–Kara Seas in both autumn and the following winter (figures not shown). Thus, the persistence of SIC anomalies in the eastern Arctic differs between the interannual and the interdecadal time scales.

It is worthwhile to note that many previous studies used September SIC indices rather than autumn SIC. Hence, we perform a correlation analysis of monthly Arctic SIC from September to November with respect to autumn SIC indices defined in the present study. The results show that on the interannual time scale, autumn ESIC and WSIC have significant correlation with monthly eastern and western Arctic SIC, respectively, during the entire autumn (figures not shown). In other words, the SIC anomalies are persistent from September to November. Thus, we can use the autumn SIC index instead of the September SIC index.

5. Impacts of the Arctic sea ice concentration anomalies

In the following analysis, we explore autumn and winter climate and atmospheric circulation anomalies associated with autumn SIC anomalies. For this purpose, a regression analysis is performed against the autumn WSIC and ESIC indices. We analyze first the concurrent anomalies in autumn and then the delayed response in winter. Again, we focus on interannual anomalies in the analysis. For comparison, a parallel analysis has been performed with respect to September WSIC and ESIC indices. The obtained temperature and circulation anomalies in autumn and winter resemble those with respect to autumn WSIC and ESIC indices.

a. Autumn

When the WSIC index is positive, significant negative temperature anomalies are located over the western Arctic, northeastern Russia, and eastern North Pacific, and significant positive temperature anomalies are seen over central North America (Fig. 6a). When the ESIC index is positive, significant cooling occurs over the eastern Arctic, subtropical western North Pacific, the southeastern coast of North America, and the eastern North Atlantic, and significant warming is seen over central Europe and the central North Atlantic (Fig. 6b). The obtained temperature anomalies in high latitudes are similar to those shown by Francis et al. (2009) based on a composite analysis of linear-trend-removed data (see their Fig. 2c).

Fig. 6.
Fig. 6.

Autumn air temperature anomalies at sigma level 0.995 obtained by regression on autumn (a) WSIC and (b) ESIC for the period 1979–2011. The contour interval is 0.2°C. The shaded areas denote anomalies that are significant at the 95% confidence level according to the Student’s t test. The outer latitude circle is 20°N.

Citation: Journal of Climate 27, 14; 10.1175/JCLI-D-13-00731.1

The distribution of autumn SLP and 500-hPa geopotential height anomalies corresponding to the ESIC and WSIC indices are shown in Fig. 7. When the WSIC index is positive, significant negative SLP anomalies extend from Bering Strait to northwestern Canada and positive anomalies occur over the midlatitudes of the North Pacific (Fig. 7a). When the ESIC index is positive, negative SLP anomalies dominate northwestern Europe and around the Beaufort Sea, and positive anomalies are observed over eastern Canada and from eastern Russia to the Aleutian Islands (Fig. 7b). The distribution of the regressed SLP anomalies resembles that obtained by Jaiser et al. (2012, see their Fig. 3b), but displays both similarity and difference compared to that obtained by Francis et al. (2009, see their Fig. 2d).

Fig. 7.
Fig. 7.

(a),(b) Autumn SLP and (c),(d) 500-hPa geopotential height and corresponding wave activity flux anomalies obtained by regression on autumn (a),(c) WSIC and (b),(d) ESIC for the period 1979–2011. The contour interval is 0.3 hPa in (a),(b) and 4 m in (c),(d). The shaded areas denote anomalies significant at the 95% confidence level according to a Student’s t test. The scale for the wave activity flux is shown at the top right of the panels. The outer latitude circle is 20°N.

Citation: Journal of Climate 27, 14; 10.1175/JCLI-D-13-00731.1

The autumn temperature anomalies have good correspondence with the sea ice and circulation anomalies. Lower temperature over the western Arctic and eastern Arctic corresponds to larger SIC in western Arctic and the eastern Arctic, respectively. This can be attributed to the reduction in the heat exchange from the ocean to the atmosphere above. The percent variance of local temperature anomalies explained by WSIC is about 10%–30% and that by ESIC is about 20%–50% (figures not shown). The distribution of temperature anomalies with respect to that of SLP anomalies indicates the contribution of anomalous wind advection. Positive temperature anomalies tend to occur east of an anomalous lows or west of anomalous highs where anomalous southerlies are expected, such as central North America (Figs. 6a and 7a) and central Europe (Figs. 6b and 7b). Negative temperature anomalies tend to appear west of anomalous lows or east of anomalous highs where anomalous northerlies are expected, such as northeastern Russia, eastern North Pacific (Figs. 6a and 7a), eastern North Atlantic, the southeast coast of North America, and the subtropical western North Pacific (Figs. 6b and 7b).

The 500-hPa geopotential height anomalies show patterns similar to those of SLP anomalies (Figs. 7a,b versus Figs. 7c,d). The similarity indicates a barotropic structure of the atmospheric response in autumn. This result is different from some previous studies. Honda et al. (2009) pointed out that there is a baroclinic response in November corresponding to SIC anomalies from the Barents to East Siberian Seas in the preceding September. Jaiser et al. (2012) showed that low SIC in late summer (August–September) triggers a baroclinic response in fall. In fact, the vertical structure of autumn atmospheric response shown by Jaiser et al. is location dependent. The response shows a barotropic structure over the North Atlantic and high latitudes of the North Pacific, but a baroclinic structure over Russia and eastern Arctic (see their Figs. 3b,f).

One way for the sea ice anomalies to induce atmospheric changes is through modulating the stability in the lower troposphere (Francis et al. 2009; Jaiser et al. 2012). The vertical profiles of temperature anomalies (Fig. 8) display a notable difference between the impact of western and eastern SIC anomalies on local atmospheric stability. When the WSIC index is high, negative temperature anomalies do not show a large difference among pressure levels in the lower to middle troposphere. This suggests that autumn SIC anomalies in the western Arctic have little effect on local atmospheric stability, but their influence on air temperature has a large vertical extension. When the ESIC index is high, the cooling is large near the surface, the temperature anomalies decrease quickly with the altitude and become weak above 900 hPa. In comparison, the vertical extent of the influence of eastern Arctic SIC anomalies is shallower, but they have a larger impact on local atmospheric stability than western Arctic SIC anomalies.

Fig. 8.
Fig. 8.

Profiles of autumn temperature anomalies (°C) averaged over the WSIC domain (solid) and ESIC domain (dashed) obtained by regression on the WSIC and ESIC index for the period 1979–2011.

Citation: Journal of Climate 27, 14; 10.1175/JCLI-D-13-00731.1

The wave activity flux at 500 hPa points eastward or southeastward over western Eurasia (Figs. 7c,d). This suggests a downstream propagation of circulation signals from northwestern Europe to Asia. This agrees with Honda et al. (2009) who indicated that anomalous heating associated with Barents–Kara Sea SIC anomalies in late autumn excited a stationary Rossby wave train propagating southeastward over Eurasia.

b. Winter

Compared to autumn temperature anomalies (Fig. 6), winter temperature anomalies are larger and more organized with main signals located over the Eurasian continent (Fig. 9). When the WSIC index is high, significant cooling is seen over the mid- to high latitudes of Eurasia and warming is located around the Okhotsk Sea (Fig. 9a). When the ESIC index is high, the temperature anomalies in the Eastern Hemisphere display a similar spatial pattern but with opposite signs (Fig. 9b). In comparison, the anomalies over the Okhotsk Sea are not obvious and there are negative anomalies over tropical central North Pacific in high ESIC years. The percent variance of Eurasian midlatitude temperature anomalies explained by WSIC is about 30%–40% and that by ESIC is about 40%–60% (figures not shown). Similar spatial pattern of winter temperature anomalies over the Eurasian continent corresponding to ESIC were obtained by previous studies (e.g., Honda et al. 2009; Wu et al. 2011). The anomalies in the Arctic are small corresponding to both indices. The results demonstrate that autumn Arctic SIC anomalies have a close relationship to the following winter climate over the Eurasian land. Note that the temperature anomalies over East Asia are confined to north of 20°N. This confirms that the Arctic sea ice anomalies mainly affect the northern mode of EAWM and have little impact on the southern mode.

Fig. 9.
Fig. 9.

As in Fig. 6, but for winter air temperature anomalies.

Citation: Journal of Climate 27, 14; 10.1175/JCLI-D-13-00731.1

When the WSIC index is high, significant positive SLP anomalies are observed over Siberia and the northwestern North Atlantic, and negative SLP anomalies occur over the western North Pacific (Fig. 10a). When the ESIC index is high, negative SLP anomalies cover most of the Eurasian continent and positive SLP anomalies cover the western North Pacific (Fig. 10b). The distribution of winter SLP anomalies over Eurasia corresponding to ESIC is consistent with previous studies (Wu et al. 2011; Jaiser et al. 2012). The correlation coefficient of the winter Siberian high with the autumn WSIC and ESIC is 0.52 and −0.60, respectively, both of which are significant at the 95% confidence level according to the Student’s t test. Here, the Siberian high intensity is represented using regionally averaged winter SLP anomaly over the region of 40°–60°N, 80°–120°E, following Wu and Wang (2002). These results indicate that the WSIC and ESIC are closely associated with the Siberian high variability.

Fig. 10.
Fig. 10.

As in Fig. 7, but for (a),(b) winter SLP and (c),(d) 500-hPa geopotential height and correspondent wave activity flux anomalies.

Citation: Journal of Climate 27, 14; 10.1175/JCLI-D-13-00731.1

The temperature anomalies have good correspondence with the SLP anomalies in winter. Corresponding to high WSIC, a lower temperature over the midlatitudes of Asia is associated with an anomalous high (Fig. 9a versus Fig. 10a). Anomalous northerlies in the eastern part of anomalous high reduce temperature through anomalous cold advection. Corresponding to high ESIC, higher temperature over Eurasia correspond to an anomalous low (Fig. 9b versus Fig. 10b). Anomalous southerlies in the eastern part of an anomalous low induce anomalous warm advection, leading to an increase in temperature.

The SLP and temperature anomalies in winter are much larger and more widespread than those in autumn. This is likely related to the seasonal change in mean state. The mean temperature gradient over Eurasian land increases largely from autumn to winter. As such, anomalous advection induces larger temperature anomalies that in turn lead to larger SLP anomalies in winter compared to autumn.

When the WSIC index is high, significant negative 500-hPa geopotential height anomalies are confined to the midlatitudes of East Asia (Fig. 10c). This indicates a deeper than normal East Asian trough. Significant positive height anomalies are seen over the western North Atlantic and around the Barents–Kara Seas and north of the Okhotsk Sea (Fig. 10c). The height anomalies over Asia display a north–south opposite pattern. When the ESIC index is high, positive height anomalies extend over midlatitudes of Asia and North Pacific (Fig. 10d). This indicates a weaker than normal East Asian trough. Negative height anomalies are observed over northwestern Europe. The spatial pattern of 500-hPa height anomalies in the Eastern Hemisphere corresponding to ESIC resembles that obtained by Jaiser et al. (2012, see their Fig. 9f) and Wu et al. (2011, see their Fig. 3c), but differences can be noticed in the location of large anomaly regions. The wave activity flux clearly indicates a southeastward extension of height anomalies from northwestern Eurasia to central and eastern Asia (Figs. 10c,d).

Comparison of the distribution of SLP (Figs. 10a,b) and 500-hPa geopotential height anomalies (Figs. 10c,d) indicates that the vertical structure is location dependent. The response of the winter atmosphere to autumn WSIC or ESIC shows a baroclinic structure over the midlatitudes of Asia, but a barotropic structure in other regions, including northern Europe, the North Pacific, and the North Atlantic. This regional dependence is due to the effect of temperature anomalies. Large temperature anomalies over midlatitude Asia lead to a baroclinic structure. In other regions where temperature anomalies are small, barotropic structure is apparent. The baroclinic nature of the atmospheric response over East Asia in December is simulated by Honda et al. (2009). Jaiser et al. (2012) indicated that low SIC in late summer (August–September) triggers a barotropic response in the following winter. In fact, comparison of their Figs. 9b and 9f shows that the vertical structure is not uniform in the entire domain. The atmospheric response is barotropic over the Arctic region and western Eurasia, whereas the response is baroclinic over the midlatitudes of East Asia.

To further illustrate the vertical structure of the atmospheric response over Asia, we show in Figs. 1112 height–latitude cross sections of temperature and geopotential height anomalies averaged over the longitudinal band 60°–120°E in winter regressed on the two SIC indices. When the WSIC index is high, the temperature anomalies are mostly confined below 300 hPa to the north of 20°N, with lower (higher) temperature over the midlatitudes (high latitudes) (Fig. 11a). The magnitude of anomalies decreases with altitude. When the ESIC index is high, the temperature anomalies show opposite signs between the midlatitudes and high latitudes in the lower to middle troposphere (Fig. 11b). Different from WSIC, the temperature anomalies in the upper troposphere are larger with the signs opposite to those in the lower troposphere.

Fig. 11.
Fig. 11.

Height–latitude cross sections of winter temperature anomalies (°C) averaged over the longitudinal band of 60°–120°E obtained by regression on the (a) WSIC and (b) ESIC for the period 1979–2011. The contour interval is 0.1°C and the zero line is omitted.

Citation: Journal of Climate 27, 14; 10.1175/JCLI-D-13-00731.1

Fig. 12.
Fig. 12.

As in Fig. 11, but for winter geopotential height anomalies with the contour interval of 3 m.

Citation: Journal of Climate 27, 14; 10.1175/JCLI-D-13-00731.1

The geopotential height anomalies display obvious different vertical distribution between mid- and high latitudes. Corresponding to high WSIC, positive height anomalies are observed north of 60°N in the entire troposphere with the magnitude of anomalies increasing with altitude (Fig. 12a). In contrast, the height anomalies between 30° and 60°N display a clear change of sign from positive near the surface to negative above 850 hPa (Fig. 12a). The height anomalies are opposite between the midlatitudes and the high latitudes above 700 hPa. Thus, the response features a baroclinic structure in the midlatitudes of Asia and a barotropic structure in the high latitudes of Asia, respectively. This change in the vertical structure is thermodynamically consistent with temperature anomalies in Fig. 11a. The contrast of the vertical structure between midlatitudes and high latitudes is obvious in geopotential height anomalies corresponding to high ESIC as well (Fig. 12b). In comparison, two differences can be noted. One is the latitude that demarcates the barotropic and baroclinic structure. It is located around 50°N in the ESIC case, 10°N southward compared to that in the WSIC case. The other difference is the obvious decrease of the magnitude of height anomalies with altitude in the upper troposphere in the ESIC case, which is not seen in the WSIC case. This difference is related to the difference in temperature anomalies in the upper troposphere in Fig. 11.

The distribution of zonal wind anomalies (Fig. 13) has a good correspondence with that of height anomalies (Fig. 12). When the WSIC index is high, anomalous westerly and easterly winds are observed over the midlatitudes and high latitudes of East Asia, respectively (Fig. 13a). This indicates a southward shift of the westerly jet over East Asia. Coupled with a deeper East Asian trough, this is in favor of advection and intrusion of more cold air to the midlatitudes and leads to colder winters over the midlatitudes of East Asia. When the ESIC index is high, opposite zonal wind anomalies are observed over East Asia (Fig. 13b). As such, the westerly jet over East Asia shifts northward. Coupled with a weaker East Asian trough, this reduces the southward advection of cold air from high latitudes, leading to warmer winters over the midlatitudes of East Asia.

Fig. 13.
Fig. 13.

As in Fig. 11, but for winter zonal wind anomalies with the contour interval of 0.3 m s−1.

Citation: Journal of Climate 27, 14; 10.1175/JCLI-D-13-00731.1

Comparing the winter anomalies associated with WSIC and ESIC (Figs. 913), the distribution over the Eurasian continent displays a similar spatial pattern albeit with opposite signs and some differences in the location of anomalies. The similarity of the spatial pattern is consistent with the significant correlation of both WSIC and ESIC with the N index, although the correlation coefficient between WSIC and ESIC is not large. The opposite anomalies are likely to a large extent contributed from the period 2004–08 when both WSIC and ESIC indices display large anomalies. Both WSIC and ESIC are closely related to the Siberian high, the East Asian trough, and the East Asian westerly jet. Since the northern mode of EAWM is associated with these mid- to high-latitude circulation systems (Chen et al. 2014), we can infer that these systems play important roles in connecting the autumn Arctic sea ice anomalies and the northern mode of the EAWM variability.

Wu et al. (2011) pointed out that there are SST anomalies concurrent with the Arctic SIC anomalies and these SST anomalies may contribute to the circulation anomalies as derived based on regression against Arctic sea ice. To examine whether this is the case, we show in Fig. 14 winter SST anomalies regressed on the WSIC and ESIC indices. There is a weak signal in the winter SST field, indicating that autumn Arctic SIC anomalies have no concurrent SST anomalies. This excludes the contribution of SST anomalies to the circulation pattern discussed above. This result differs from that of Wu et al. (2011). The discrepancy is likely due to the difference in the time scale. Wu et al. (2011) did not separate the interannual and interdecadal components, whereas the present study focuses on the interannual time scale. If we did not exclude the interdecadal variations, the results are distinct from those shown in Fig. 14. Figure 15 shows the regressed winter SST anomalies on the WSIC and ESIC indices without time filtering applied to the variables. Corresponding to the WSIC, there are still no significant SST anomalies except for the Aleutian region (Fig. 15a). However, corresponding to the ESIC, significant SST anomalies appear in the North Pacific and North Atlantic. Positive anomalies are seen in the eastern North Pacific and along the west coast of North America. Negative anomalies are observed in the western North Pacific, the tropics, and high latitudes of the North Atlantic as well as the Arctic Ocean to the east of Greenland (Fig. 15b). The spatial pattern of SST anomalies resembles that obtained by Wu et al. (2011, see their Fig. 3f). This indicates that the close relationship between autumn eastern Arctic SIC and the following winter SST is attributed to the interdecadal variability in the ESIC and SST.

Fig. 14.
Fig. 14.

As in Fig. 6, but for winter SST anomalies. The contour interval is 0.1°C. The outer latitude circle is 0°.

Citation: Journal of Climate 27, 14; 10.1175/JCLI-D-13-00731.1

Fig. 15.
Fig. 15.

As in Fig. 6, but for original winter SST anomalies. The contour interval is 0.1°C. The outer latitude circle is 0°.

Citation: Journal of Climate 27, 14; 10.1175/JCLI-D-13-00731.1

The declining Arctic sea ice plays a critical role in driving heavy snowfall over Europe during winter (e.g., Liu et al. 2012b) and the snow cover over the Eurasian continent may affect the EAWM variability (e.g., Gong et al. 2003; Jhun and Lee 2004; Wang et al. 2010). So, the snow cover change may serve as a communicator for the influence of autumn Arctic sea ice anomalies on the EAWM variability. Figure 16 shows winter snow cover anomalies regressed on the WSIC and ESIC indices. Corresponding to positive WSIC, significant positive snow cover anomalies are seen over Europe and eastern China (Fig. 16a). A similar spatial pattern of snow cover anomalies is observed corresponding to the ESIC but with opposite signs (Fig. 16b). The distribution of snow cover anomalies over Eurasia corresponding to the ESIC is consistent with previous studies (Honda et al. 2009; Orsolini et al. 2012).

Fig. 16.
Fig. 16.

Winter snow cover frequency anomalies obtained by regression on autumn (a) WSIC and (b) ESIC for the period 1979–2010.

Citation: Journal of Climate 27, 14; 10.1175/JCLI-D-13-00731.1

The snow cover anomalies appear as responses to circulation and temperature changes. When the WSIC (ESIC) is high (low), above-normal snow cover over Europe and eastern China is associated with anomalous lower surface temperature (Fig. 9) and anomalous northerly winds (Figs. 10a,b). In turn, the snow cover anomalies may affect surface temperature through modifying the surface albedo and reducing the solar radiation absorption. It is difficult to separate out the snow cover effect based on the observations. Further studies are needed to address this issue.

c. Plausible processes connecting autumn Arctic sea ice to winter Eurasian climate

Analyses in section 5b demonstrate obvious signals of winter Asian climate in preceding autumn Arctic sea ice anomalies. This suggests potential predictability of the EAWM from the Arctic sea ice anomalies. One issue is what are the plausible physical processes that link the autumn Arctic sea ice to winter Asian climate anomalies.

Previous studies have proposed processes for the influence of Arctic sea ice on winter Eurasian climate. Honda et al. (2009) suggested the role of stationary Rossby waves. Anomalously low sea ice cover in late autumn over the Barents–Kara Seas generates a stationary Rossby wave through anomalous turbulent heat fluxes. The southeastward propagation of wave activity flux leads to weakening of storm-track activity along 60°N in winter. This yields anomalous cold advection southwestward, contributing to the intensification of the Siberian high that brings more cold air masses to far east Asia. Wu et al. (2011) showed that more sea ice concentration over the Barents–Kara Seas in autumn–winter induces an anomalous low in winter over northern Eurasia, weakening the Siberian high and intensifying westerlies over the mid- to high latitudes of Eurasia. The strengthened westerlies are unfavorable for the southward intrusion of cold air from high latitudes, leading to a higher surface temperature over the mid- to high latitudes of Eurasia. Wu et al. (2011) also suggested a contribution of air temperature cooling over the Arctic and associated north–south thermal gradient between the Arctic and the mid- to high latitudes of Eurasia to the strengthened westerlies over northern Eurasia. Jaiser et al. (2012) indicated that the increase of heat release over the Arctic Ocean to the atmosphere accompanying less Arctic sea ice in August–September reduces the atmospheric stability. This leads to an earlier onset of baroclinic instability that modulates the meridional heat flux and temperature gradient in autumn, which, in turn, induces a northward shift of storm activity in connection with planetary waves in winter.

According to previous studies, there are two plausible pathways by which autumn Arctic SIC anomalies affect winter Asian circulation. One way is that anomalous autumn SIC-induced thermal state changes persist into winter and then influence atmospheric circulation. The other way is that anomalous autumn SIC-induced thermal state leads to atmospheric circulation around the Arctic and then the circulation change extends to Asia in winter through atmospheric processes, such as the wave activity flux. Comparing Figs. 6 and 9, the temperature anomalies over the Arctic Ocean display differences between winter and autumn. This indicates that the local thermodynamic effects of Arctic sea ice anomalies cannot persist. As such, the first pathway appears not to be working.

Based on autumn and winter temperature and circulation anomalies, we propose the following processes connecting autumn Arctic SIC anomalies to winter Asian climate anomalies and take the ESIC as an example in describing the processes. A similar argument works for the processes of influence of WSIC except for opposite anomalies. Following more eastern Arctic SIC in autumn, surface air temperature decreases over the eastern Arctic (Fig. 6b) due to reduced heat fluxes from ocean to atmosphere. In response, SLP increases over the eastern Arctic and coastal Russia due to thermodynamic effects (Fig. 7b). This leads to a shift of the polar high to the side of Russia and Bering Strait. Correspondingly, SLP decreases on the other side of the Arctic, that is, northwestern Europe (Fig. 7b). As temperature anomalies are small over northwestern Europe (Fig. 6b), a barotropic structure follows, and the 500-hPa height is lower than normal as well (Fig. 7d). Owing to the southeastward propagation of wave activity flux, negative height anomalies extend eastward and positive height anomalies form to the southeast over midlatitude Asia in winter (Fig. 10d). The meridional gradient of height anomalies weakens the westerly jet stream over East Asia (Fig. 13b) and results in a northward shift of storm activity. This leads to a weakened Siberian high (Fig. 10b) and a warming of midlatitude Asia (Fig. 9b).

6. Summary and discussion

The present study investigated the impacts of autumn Arctic sea ice concentration anomalies on the two modes of East Asian winter monsoon variability, and focuses on the interannual time scale. It is found that both western and eastern Arctic SIC anomalies in autumn can affect the northern mode of EAWM significantly and the relation between the northern mode and SIC anomalies in western and eastern Arctic regions tends to be opposite. However, the southern mode of the EAWM variability does not appear to be related to the Arctic sea ice anomalies except in small regions. The results indicate that there is some predictability of the northern mode of EAWM variability from the previous autumn Arctic sea ice anomalies, but the source of predictability of the southern mode from the Arctic sea ice is likely low. Based on the distribution of the correlation of autumn SIC with an index for the northern mode of EAWM (N index), two SIC indices have been defined to measure the variations of autumn SIC in the western and eastern Arctic, denoted as WSIC and ESIC, respectively. Different from previous studies, the present study analyzes the impacts of not only eastern Arctic SIC anomalies, but also western Arctic SIC anomalies that have received little attention before.

On the interannual time scale, the variations of autumn SIC in eastern and western Arctic regions tend to be independent of each other during the analysis period. The SIC anomalies in the western Arctic display a tendency of persistence from autumn to winter, whereas those in the eastern Arctic do not show obvious persistence. The coherent features of SIC anomalies in the eastern Arctic from autumn to winter obtained by Wu et al. (2011) are attributed to the contribution of the interdecadal component of SIC variations.

The Arctic SIC anomalies can exert influences locally as well as in remote regions in autumn. Local surface air temperature decreases in response to more sea ice. The local temperature response to western Arctic SIC anomalies has a much deeper vertical extension than that to eastern Arctic SIC anomalies, with the former extending to the middle troposphere and the latter being confined to the lower troposphere. As such, eastern Arctic SIC anomalies have a much larger influence on the stability of the lower troposphere than western Arctic SIC anomalies. The distribution of autumn atmospheric circulation anomalies corresponding to the WSIC and ESIC are different from each other, and there is no obvious signal over most of the Eurasian continent. The atmospheric response displays a barotropic structure over most of the North Hemisphere corresponding to both western and eastern Arctic SIC anomalies.

Widespread and large surface air temperature anomalies are observed in winter over the mid- to high latitudes of the Eurasian continent corresponding to autumn Arctic SIC anomalies. The winter temperature anomalies are small in other regions. The anomalous circulation in winter corresponding to the WSIC and ESIC displays a similar spatial pattern but with opposite signs. Both western and eastern Arctic SIC anomalies in autumn are followed by pronounced changes in circulation systems over the Eurasian continent, including the Siberian high, the East Asian trough, and the East Asian westerly jet stream. Thus, these systems may play important roles in connecting autumn Arctic SIC anomalies to the northern mode of EAWM and associated climate anomalies in winter over East Asia. In winter, the atmospheric response shows a baroclinic structure over the midlatitudes of Asia and a barotropic structure in other regions, such as northern Europe, the North Pacific, and the North Atlantic.

The autumn Arctic SIC anomalies have no large concurrent SST anomalies on the interannual time scale. This indicates that the Arctic sea ice anomalies impact the winter climate over East Asia independent of SST forcing. The winter SST anomalies following September eastern Arctic SIC changes obtained by Wu et al. (2011) are attributed to the interdecadal component of variations in the Arctic SIC and SST. Corresponding to autumn Arctic SIC anomalies, obvious snow cover anomalies occur in winter over the Eurasian continent. The snow cover anomaly pattern appears to result from atmospheric circulation and surface temperature changes. How the snow cover affects the northern mode of EAWM is unclear and remains to be investigated.

The present analysis identifies obvious anomalies in temperature and atmospheric circulation in both autumn and winter following autumn Arctic SIC anomalies. The regions where temperature and circulation anomalies are observed display prominent differences between autumn and winter. Temperature and circulation anomalies are weak over Asia in autumn, whereas large anomalies occur over Asia in winter. This is likely due to a large meridional gradient of mean temperature in winter over the mid- and high latitudes of Asia under which anomalous winds induce large temperature anomalies. The connection from autumn Arctic SIC anomalies to winter Asian circulation and temperature anomalies appears to be through an anomalous pressure pattern around the Arctic in autumn via the thermodynamic effect of SIC anomalies and a downstream extension of circulation anomalies to Asia in winter through the propagation of wave activity flux. Further studies are needed to understand the mechanism for the influence of autumn Arctic SIC anomalies on the winter climate over Asia.

Acknowledgments

We appreciate comments of three anonymous reviewers. The study is supported by National Basic Research Program of China Grant 2014CB953902 and National Natural Science Foundations of China Grants 41275081 and 41228006. RW acknowledges the support of the Hong Kong Research Grants Council Grant CUHK403612.

REFERENCES

  • Alexander, M. A., , U. S. Bhatt, , J. E. Walsh, , M. S. Timlin, , J. S. Miller, , and J. D. Scott, 2004: The atmospheric response to realistic Arctic sea ice anomalies in an AGCM during winter. J. Climate, 17, 890905, doi:10.1175/1520-0442(2004)017<0890:TARTRA>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Brodzik, M., , and R. Armstrong, 2013: Northern Hemisphere EASE-Grid 2.0 weekly snow cover and sea ice extent version 4. National Snow and Ice Data Center, Boulder, CO, digital media. [Available online at https://nsidc.org/data/docs/daac/nsidc0046_nh_ease_snow_seaice.gd.html.]

  • Chen, Z., , R. Wu, , and W. Chen, 2014: Distinguishing interannual variations of the northern and southern modes of the East Asian winter monsoon. J. Climate, 27, 835851, doi:10.1175/JCLI-D-13-00314.1.

    • Search Google Scholar
    • Export Citation
  • Comiso, J. C., , C. L. Parkinson, , R. Gersten, , and L. Stock, 2008: Accelerated decline in the Arctic sea ice cover. Geophys. Res. Lett., 35, L01703, doi:10.1029/2007GL031972.

    • Search Google Scholar
    • Export Citation
  • Deser, C., , G. Magnusdottir, , R. Saravanan, , and A. Philips, 2004: The effects of North Atlantic SST and sea ice anomalies on the winter circulation in CCM3. Part II: Direct and indirect components of the response. J. Climate, 17, 877889, doi:10.1175/1520-0442(2004)017<0877:TEONAS>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Francis, J. A., , W. Chan, , D. J. Leathers, , J. R. Miller, , and D. E. Veron, 2009: Winter Northern Hemisphere weather patterns remember summer Arctic sea-ice extent. Geophys. Res. Lett., 36, L07503, doi:10.1029/2009GL037274.

    • Search Google Scholar
    • Export Citation
  • Gong, G., , D. Entekhabi, , and J. Cohen, 2003: Modeled Northern Hemisphere winter climate response to realistic snow anomalies. J. Climate, 16, 39173931, doi:10.1175/1520-0442(2003)016<3917:MNHWCR>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • He, S.-P., , and H.-J. Wang, 2013: An integrated East Asian winter monsoon index and its interannual variability (in Chinese). Chin. J. Atmos. Sci., 36 (3), 523538.

    • Search Google Scholar
    • Export Citation
  • Honda, M., , J. Inoue, , and S. Yamane, 2009: Influence of low Arctic sea-ice minima on anomalously cold Eurasian winters. Geophys. Res. Lett., 36, L08707, doi:10.1029/2008GL037079.

    • Search Google Scholar
    • Export Citation
  • Hopsch, S., , J. Cohen, , and K. Dethloff, 2012: Analysis of a link between fall Arctic sea ice concentration and atmospheric patterns in the following winter. Tellus, 64A, 18624, doi:10.3402/tellusa.v64i0.18624.

    • Search Google Scholar
    • Export Citation
  • Inoue, J., , M. E. Hori, , and K. Takaya, 2012: The role of Barents Sea ice in the wintertime cyclone track and emergence of a warm-Arctic cold-Siberian anomaly. J. Climate, 25, 25612568, doi:10.1175/JCLI-D-11-00449.1.

    • Search Google Scholar
    • Export Citation
  • Jaiser, R., , K. Dethloff, , D. Handorf, , A. Rinke, , and J. Cohen, 2012: Impact of sea ice cover changes on the Northern Hemisphere atmospheric winter circulation. Tellus, 64A, 11595, doi:10.3402/tellusa.v64i0.11595.

    • Search Google Scholar
    • Export Citation
  • Jhun, J.-G., , and E.-J. Lee, 2004: A new East Asian winter monsoon index and associated characteristics of the winter monsoon. J. Climate, 17, 711726, doi:10.1175/1520-0442(2004)017<0711:ANEAWM>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Kalnay, E., and Coauthors, 1996: The NCEP/NCAR 40-Year Reanalysis Project. Bull. Amer. Meteor. Soc., 77, 437471, doi:10.1175/1520-0477(1996)077<0437:TNYRP>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Kang, L., , W. Chen, , L. Wang, , and L. Chen, 2009: Interannual variations of winter temperature in China and their relationship with the atmospheric circulation and sea surface temperature (in Chinese). Climate Environ. Res., 14 (1), 4553.

    • Search Google Scholar
    • Export Citation
  • Li, F., , and H.-J. Wang, 2013: Relationship between Bering Sea ice cover and East Asian winter monsoon year-to-year variations. Adv. Atmos. Sci., 30, 4856, doi:10.1007/s00376-012-2071-2.

    • Search Google Scholar
    • Export Citation
  • Liu, G., , L.-R. Ji, , S.-Q. Sun, , and Y.-F. Xin, 2012a: Low- and mid-high latitude components of the East Asian winter monsoon and their reflecting variations in winter climate over eastern China. Atmos. Oceanic Sci. Lett., 5, 195200.

    • Search Google Scholar
    • Export Citation
  • Liu, J., , J. A. Curry, , H.-J. Wang, , M. R. Song, , and R. M. Horton, 2012b: Impact of declining Arctic sea ice on winter snowfall. Proc. Natl. Acad. Sci. USA, 109, 40744079, doi:10.1073/pnas.1114910109.

    • Search Google Scholar
    • Export Citation
  • Orsolini, Y. J., , R. Senan, , R. E. Benestad, , and A. Melsom, 2012: Autumn atmospheric response to the 2007 low Arctic sea ice extent in coupled ocean-atmosphere hindcasts. Climate Dyn., 38, 24372448, doi:10.1007/s00382-011-1169-z.

    • Search Google Scholar
    • Export Citation
  • Rayner, N. A., , D. E. Parker, , E. B. Horton, , C. K. Folland, , L. V. Alexander, , D. P. Rowell, , E. C. Kent, , and A. Kaplan, 2003: Global analyses of sea surface temperature, sea ice, and night marine air temperature since the late nineteenth century. J. Geophys. Res., 108, 4407, doi:10.1029/2002JD002670.

    • Search Google Scholar
    • Export Citation
  • Smith, T. M., , R. W. Reynolds, , T. C. Peterson, , and J. Lawrimore, 2008: Improvements to NOAA’s historical merged land–ocean surface temperature analysis (1880–2006). J. Climate, 21, 22832296, doi:10.1175/2007JCLI2100.1.

    • Search Google Scholar
    • Export Citation
  • Wang, B., , Z. Wu, , C. P. Chang, , J. Liu, , J. Li, , and T. Zhou, 2010: Another look at interannual-to-interdecadal variations of the East Asian winter monsoon: The northern and southern temperature modes. J. Climate, 23, 14951512, doi:10.1175/2009JCLI3243.1.

    • Search Google Scholar
    • Export Citation
  • Wang, L., , and W. Chen, 2010: How well do existing indices measure the strength of the East Asian winter monsoon? Adv. Atmos. Sci., 27, 855870, doi:10.1007/s00376-009-9094-3.

    • Search Google Scholar
    • Export Citation
  • Watanabe, M., , and T. Nitta, 1999: Decadal change in the atmospheric circulation and associated surface climate variations in the Northern Hemispheric winter. J. Climate, 12, 494510, doi:10.1175/1520-0442(1999)012<0494:DCITAC>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Wu, B.-Y., , and J. Wang, 2002: Winter Arctic Oscillation, Siberian High and East Asian winter monsoon. Geophys. Res. Lett., 29, 1897, doi:10.1029/2002GL015373.

    • Search Google Scholar
    • Export Citation
  • Wu, B.-Y., , R.-H. Huang, , and D.-Y. Gao, 1999: Impact of variations of winter sea-ice extents in the Kara/Barents Seas on winter monsoon over East Asia (in Chinese). Acta Meteor. Sin., 13, 141153.

    • Search Google Scholar
    • Export Citation
  • Wu, B.-Y., , R.-H. Zhang, , and R. Arrigo, 2006: Distinct modes of the East Asian winter monsoon. Mon. Wea. Rev., 134, 21652179, doi:10.1175/MWR3150.1.

    • Search Google Scholar
    • Export Citation
  • Wu, B.-Y., , J.-Z. Su, , and R.-H. Zhang, 2011: Effects of autumn-winter Arctic sea ice on winter Siberian High. Chin. Sci. Bull., 56, 32203228, doi:10.1007/s11434-011-4696-4.

    • Search Google Scholar
    • Export Citation
Save